Semi-structured interviewing balances structure with flexibility: you prepare a guide, but you follow the participant’s meaning rather than a rigid script. Design choices should match your methodology, data needs, and ethical constraints.
Types of interviews
Structured interviews use fixed wording and order. They maximize comparability but can constrain emergent meaning; uncommon in classic grounded theory (GT) except for specific comparison needs.
Semi-structured interviews use a guide with core topics and optional probes. They are the default for most qualitative work because they support depth and adaptation.
Unstructured interviews resemble guided conversations with minimal predetermined questions. Useful for early exploration when you want maximum openness—often paired with strong memoing and later more focused sessions.
Developing an interview guide
List what you need to learn (processes, meanings, contexts), not what you assume is true.
관련 스킬
Anchor in purpose.
Order for rapport. Begin with non-threatening, descriptive items; place sensitive topics after trust is established unless safety protocols require otherwise.
Use open stems. Prefer “Tell me about…,” “Walk me through…,” “What happened when…?” over yes/no or leading frames.
Build flexibility. For each topic, note 2–4 optional probes you can use if the participant’s answer is thin or tangential.
Pilot and revise. Run the guide with a volunteer or colleague; time the flow, remove redundant items, and fix confusing wording.
Question types (techniques)
Grand tour: Broad invitation to narrate a domain (e.g., “Describe a typical day when X happens.”).
Mini tour: Zoom into a specific episode (e.g., “Pick one instance and walk me through it moment by moment.”).
Probing: Clarify, elaborate, or specify without steering (e.g., “What do you mean by…?”, “What happened next?”, “Who was involved?”).
Follow-up: Return to earlier content when new threads appear (e.g., “Earlier you said Y—how does that connect to what you just described?”).
Pilot testing
Pilot at least one complete interview when feasible. Record what questions misfire, which order feels awkward, and how long sections actually take. Update the guide before “real” data collection begins.
Rapport and interactional quality
Explain the study plainly; normalize pauses; avoid interrupting mid-thought; mirror the participant’s vocabulary before introducing your own; acknowledge emotion without rushing to “fix” it. For video/phone formats, test technology beforehand and have a backup plan.
Recording and transcription
Obtain consent for recording. Use high-quality audio; label files immediately. Plan transcription approach (verbatim vs intelligent verbatim), turnaround time, and how you will store files securely. Note nonverbal cues in a separate column or memo if they matter analytically.
Grounded theory considerations
GT favors open questions and following the data in situ. Treat the guide as a temporary scaffold: if a participant introduces a theoretically important incident, pursue it even if it skips ahead in your outline. Avoid importing a literature-derived codebook into question wording early on. Memo during or immediately after each interview to capture emergent ideas before they fade.
Quick checklist
Guide matches methodology (especially GT openness vs structured comparison needs).
Questions are open, non-leading, and ordered for rapport.